When the stems of dark-grown soybean [Glycine max (L.) Merr.] seedlings grew rapidly at favorable temperatures in saturating humidities, a water potential of about 0.2 MPa was induced by growth (Psi(o)-Psi(w), where Psi(o), is the water potential of the basal nonelongating tissue and Psi(w) is the water potential of the elongating tissue). If this water potential was caused by high concentrations of solute in the apoplast, as has been proposed, lowering the temperature should have little effect on the potential. On the other hand, if the water potential was caused by apoplast tensions generated by growth, then the tensions should disappear as growth is inhibited by low temperatures. We observed that the growth-induced water potential became too small to detect when growth was inhibited by temperatures as low as 13.5 degrees C. The disappearance was observed as a rise in apoplast water potential using a thermocouple psychrometer for intact plants, a rise in cell turgor using a miniature pressure probe and a decrease in apoplast tensions using a pressure chamber. The disappearance was not caused by a loss of solute from the apoplast because the tensions fully accounted for the growth-induced water potential at all temperatures. The results are consistent with the lack of solute measured directly in the apoplast solutions at high temperatures (Nonami and Boyer 1987). Therefore, it was concluded that little solute was present in the apoplast at any temperature, and the growth-induced water potential was associated mostly with a tension that moved water from the xylem and into the surrounding cells to meet the demand of cell enlargement.